Question 1

Time-series data are data collected for a

Accepted Answer

A) given individual or individuals for a number of different time-periods. 
B) given individual for one time-period. 
C) number of individuals for one time-period. 
D) number of individuals for a number of different time-periods. 
A) given individual or individuals for a number of different time-periods. 
B) given individual for one time-period. 
C) number of individuals for one time-period. 
D) number of individuals for a number of different time-periods. A

Question 2

What does it mean for a time-series to be stationary? Why is this desirable? Explain.

Accepted Answer

A time-series is stationary if the probability distribution of the dependent variable is constant over time.Stationarity is a desirable property because if the probability distribution of the dependent variable changes over time it becomes difficult to explain the variation in that variable.

Question 3

You can de-trend your data by

Accepted Answer

A) performing OLS on the de-trended data. 
B) performing weighted least squares on the initial data by diving the initial data by the estimate coefficient on the time-trend variable. 
C) separately regressing the dependent variable and each of the independent on the time trend and subtracting the estimated residuals from the initial data. 
D) including a time trend in your OLS regression. 
A) performing OLS on the de-trended data. 
B) performing weighted least squares on the initial data by diving the initial data by the estimate coefficient on the time-trend variable. 
C) separately regressing the dependent variable and each of the independent on the time trend and subtracting the estimated residuals from the initial data. 
D) including a time trend in your OLS regression. C

Question 4

Suppose you estimate a distributed lag model of U.S.Personal Consumption Expenditures (billions)on U.S.Personal Income (billions)using quarterly data for the period 1969Q1-2013Q3 and you get $$\begin{array} { l } 
\text { Consumption } _ { t } =& - 135.034 +& 0.757 \text { Income } _ { t } +& 0.163 \text { Income } _ { t - 1 } -& 0.243 \text { Income } _ { t - 2 }-& 0.158 \text { Income } _ { t - 3 } +& 0.307 \text { Income } _ { t - 4 } \
&( 11.244 )& ( 0.099 ) &( 0.153 ) & ( 0.151 )&(0.159) & (0.107) \
&&&n = 175 \
&&&R ^ { 2 } = .9994 \
\end{array}$$ 
a)What is this model called? Why would you estimate this type of function?
b)Are the lags individually significant at the 5% level?
c)What do your results in part b imply about your preferred model?

Accepted Answer

a)This is a distributed lag model and it

Question 5

What is forecasting? How do you perform it? How can you test the potential validity of your forecasts? Explain.

Accepted Answer

Forecasting is using regression results

Question 6

A structural break occurs when we see

Accepted Answer

A) an unexpected shift in time-series data. 
B) a number of outliers in cross-section data. 
C) a general upward trend over time in time-series data. 
D) an independent variable is correlated with the dependent variable but there is no theoretical justification on for the relationship. 
A) an unexpected shift in time-series data. 
B) a number of outliers in cross-section data. 
C) a general upward trend over time in time-series data. 
D) an independent variable is correlated with the dependent variable but there is no theoretical justification on for the relationship.

Question 7

You can limit the problem of potential spurious correlation by

Accepted Answer

A) estimating the regression both with and without the suspect independent variable and comparing the R-squared for the two regressions. 
B) thinking carefully about the chose model specification to rule out potential spurious correlations. 
C) including interaction terms between the suspect independent variable and the remaining independent variables and testing for the joint significance of those terms. 
D) including a dummy variable for the suspect independent variable and testing for the individual significance of that term. 
A) estimating the regression both with and without the suspect independent variable and comparing the R-squared for the two regressions. 
B) thinking carefully about the chose model specification to rule out potential spurious correlations. 
C) including interaction terms between the suspect independent variable and the remaining independent variables and testing for the joint significance of those terms. 
D) including a dummy variable for the suspect independent variable and testing for the individual significance of that term.

Question 8

Suppose you regress Ice Cream Sales on the number of children under age 10 and on binary dummy variables indicating the 2nd,3rd,and 4th quarters of the year and you get (p-values in parentheses)

$$\begin{array} { lllll } 
Ice ~\widehat {Cream}~ Sales { } _ { t } =& 17.92 +& 8.63 Under 10 _ { t } + &14.61 Q 2 _ { t } + &5.32 Q _ { t } +& 1.08 Q 4 _ { t }\
&( 0.012 )& ( 0.032 )&( 0.037 )& ( 0.124 )&( 0.179 ) \

\end{array}$$
You should conclude that,holding all other independent variables constant,ice cream sales are statistically greater in

Accepted Answer

A) Q1 than in Q4. 
B) Q1 than in Q3. 
C) Q2 than in Q1. 
D) Q1 than in Q2. 
A) Q1 than in Q4. 
B) Q1 than in Q3. 
C) Q2 than in Q1. 
D) Q1 than in Q2.

Question 9

You can test for a structural break in time-series data by

Accepted Answer

A) including a dummy variable for data after the suspected break and interaction terms between that variable and the remaining independent variables and testing for the joint significance of those terms. 
B) including dummy variables for three of the four quarters and testing for the joint significance of those variables. 
C) estimating the regression both with and without the suspect independent variable and comparing the R-squared for the two regressions. 
D) including a time-trend and testing for the individual significance of that term. 
A) including a dummy variable for data after the suspected break and interaction terms between that variable and the remaining independent variables and testing for the joint significance of those terms. 
B) including dummy variables for three of the four quarters and testing for the joint significance of those variables. 
C) estimating the regression both with and without the suspect independent variable and comparing the R-squared for the two regressions. 
D) including a time-trend and testing for the individual significance of that term.

Question 10

Suppose you estimate a static time-series model of U.S.Personal Consumption Expenditures (billions)on U.S.Personal Income (billions)using quarterly data for the period 1969Q1-2013Q3 and you get $$\begin{array} { c } 
\text { Consumption } _ { t } =& - 141.454 +& 0.822 \text { Income } _ { t } \
&( 11.244 )& ( 0.002 ) \
&n = 179 \
&R ^ { 2 } = .9993
\end{array}$$ 
a)How do you interpret the estimated sample regression function? Explain.
b)You wonder if previous values of income affect consumption.Explain how you would estimate a distributed lag model of order 2.
c)Using the model in part c,how would you test for a statistical relationship between past values of income and consumption?
d)If you thought there was a structural break in these data in 2006Q1,how would you control and test for it?

Accepted Answer

a)This is a consumption function.The int

Question 11

What is a structural break in the data? Why is it important to control for a structural break if one occurs? How can you test and control for a structural break in your data? Explain.

Accepted Answer

A structural break occurs when there is

Question 12

What is a time-trend? Why is it important to control for a time-trend if it is present in the data? How can you test and control for a time-trend in your data? Explain.

Accepted Answer

A time-trend exists when the dependent a

Question 13

A weakly dependent time series is one for which

Accepted Answer

A) the observations for a given variable become more closely related over time. 
B) the probability of distribution of the dependent variable changes over time. 
C) the probability of distribution of the dependent variable does not changes over time. 
D) the observations for a given variable become less closely related over time. 
A) the observations for a given variable become more closely related over time. 
B) the probability of distribution of the dependent variable changes over time. 
C) the probability of distribution of the dependent variable does not changes over time. 
D) the observations for a given variable become less closely related over time.

Question 14

What is spurious correlation? Why is it important to consider whether spurious correlation exists? How can you do so? Explain.

Accepted Answer

Spurious correlation occurs when a stati

Question 15

Suppose that you regress obesity rates on smoking rates for the years 1928-2011 by the model $\text { Obesity Rate } _ { t } = \beta _ { 0 } + \beta _ { 1 } \text { Smoking } ^ { \text {Rate } } + \varepsilon _ { t }$
)You have just estimated a _______ model.

Accepted Answer

A) distributed lagged 
B) time-trend 
C) seasonality 
D) static time-series 
A) distributed lagged 
B) time-trend 
C) seasonality 
D) static time-series

Question 16

What is seasonality? When might you expect seasonality to occur? Why is it important to control for a seasonality if it is present in the data? How can you test and control for a seasonality in your data? Explain.

Accepted Answer

Seasonality occurs when the dependent va

Question 17

A stationary time-series is one for which

Accepted Answer

A) the probability of distribution of the dependent variable changes over time. 
B) the probability of distribution of the dependent variable does not change over time. 
C) the probability of distribution of the independent variables changes over time. 
D) the probability of distribution of the independent variables does not change over time. 
A) the probability of distribution of the dependent variable changes over time. 
B) the probability of distribution of the dependent variable does not change over time. 
C) the probability of distribution of the independent variables changes over time. 
D) the probability of distribution of the independent variables does not change over time.

Question 18

A static time-series model _______ for the time dependence of the data.

Accepted Answer

A) is estimated by OLS with no controls 
B) includes lagged dependent variables to control 
C) includes lagged independent variables to control 
D) includes weighted residuals to control 
A) is estimated by OLS with no controls 
B) includes lagged dependent variables to control 
C) includes lagged independent variables to control 
D) includes weighted residuals to control

Question 19

Suppose you include seasonality controls in your model of U.S.Personal Consumption Expenditures (billions)on U.S.Personal Income (billions)using quarterly data for the period 1969Q1-2013Q3 and you get $$\begin{array} { c } 
\text { Consumption } _ { t } =& - 143.783 +&  0.822 \text { Income } _ { t } - & 1.782 Q 2 _ { t } + & 11.431 Q 3 _ { t } -&  0.073 Q 4 _ { t } \
& ( 15.775 ) & ( 0.002 )& ( 18.098 ) & ( 18.099 ) & ( 18.200 ) \
& &  { n } = 179 \
& & R ^ { 2 } = .9994
\end{array}$$ 
A)Explain how you would create Q2t,Q3t,and Q4t.
B)Interpret the coefficients on Q2t,Q3t,and Q4t and comment on their statistical significance.

Accepted Answer

a)To create Q2_t put a 1 for the first qua

Question 20

What are the assumptions required for OLS to be BLUE for time-series data? Write out and explain each.

Accepted Answer

TS1 - Linear in the parameters.This mean

Time-series data are data collected for a

What does it mean for a time-series to be stationary? Why is this desirable? Explain.

You can de-trend your data by

What is forecasting? How do you perform it? How can you test the potential validity of your forecasts? Explain.

A structural break occurs when we see

You can limit the problem of potential spurious correlation by

You can test for a structural break in time-series data by

What is a structural break in the data? Why is it important to control for a structural break if one occurs? How can you test and control for a structural break in your data? Explain.

What is a time-trend? Why is it important to control for a time-trend if it is present in the data? How can you test and control for a time-trend in your data? Explain.

A weakly dependent time series is one for which

What is spurious correlation? Why is it important to consider whether spurious correlation exists? How can you do so? Explain.

Suppose that you regress obesity rates on smoking rates for the years 1928-2011 by the model $\text { Obesity Rate } _ { t } = \beta _ { 0 } + \beta _ { 1 } \text { Smoking } ^ { \text {Rate } } + \varepsilon _ { t }$ )You have just estimated a _______ model.

What is seasonality? When might you expect seasonality to occur? Why is it important to control for a seasonality if it is present in the data? How can you test and control for a seasonality in your data? Explain.

A stationary time-series is one for which

A static time-series model _______ for the time dependence of the data.

What are the assumptions required for OLS to be BLUE for time-series data? Write out and explain each.

An Introduction to Econometrics and Statistical Inference

Collection and Management of Data

Summary Statistics

Simple Linear Regression

Hypothesis Testing in Linear Regression Analysis

Multiple Linear Regression Analysis

Qualitative Variables and Non-Linearities in Multiple Linear Regression Analysis

Model Selection in Multiple Linear Regression Analysis

Heteroskedasticity

Auto-Correlation

Limited Dependent Variables

Panel Data

Instrumental Variables for Simultaneous Equations, Endogenous Independent Variables, and Measurement Error

Quantile Regression, Count Data, Sample Selection Bias, and Quasi-Experimental Methods

Filters

Exam 10: Time Series Analysis

Time-series data are data collected for a

What does it mean for a time-series to be stationary? Why is this desirable? Explain.

You can de-trend your data by

What is forecasting? How do you perform it? How can you test the potential validity of your forecasts? Explain.

A structural break occurs when we see

You can limit the problem of potential spurious correlation by

You can test for a structural break in time-series data by

What is a structural break in the data? Why is it important to control for a structural break if one occurs? How can you test and control for a structural break in your data? Explain.

What is a time-trend? Why is it important to control for a time-trend if it is present in the data? How can you test and control for a time-trend in your data? Explain.

A weakly dependent time series is one for which

What is spurious correlation? Why is it important to consider whether spurious correlation exists? How can you do so? Explain.

What is seasonality? When might you expect seasonality to occur? Why is it important to control for a seasonality if it is present in the data? How can you test and control for a seasonality in your data? Explain.

A stationary time-series is one for which

A static time-series model _______ for the time dependence of the data.

What are the assumptions required for OLS to be BLUE for time-series data? Write out and explain each.

An Introduction to Econometrics and Statistical Inference

Collection and Management of Data

Summary Statistics

Simple Linear Regression

Hypothesis Testing in Linear Regression Analysis

Multiple Linear Regression Analysis

Qualitative Variables and Non-Linearities in Multiple Linear Regression Analysis

Model Selection in Multiple Linear Regression Analysis

Heteroskedasticity

Auto-Correlation

Limited Dependent Variables

Panel Data

Instrumental Variables for Simultaneous Equations, Endogenous Independent Variables, and Measurement Error

Quantile Regression, Count Data, Sample Selection Bias, and Quasi-Experimental Methods

Filters